PhD in AI before it was so hip & hyped. Synth artist & published photographer of Transylvanian Hungarian origin, based in the UK... keeping it gothic :)
@ValerioCapraro This has been crystal clear fundamental known thing to anyone who understood how LLMs, by definition, work. But the astounding hype & BS now has resulted in a surreal situation where fundamental facts need to be hammered into totally brainwashed idiots' brains.
@zoltan_the_hun_@Maks_NAFO_FELLA There are no exit polls. And it says it is estimation based on previous 5 days of polls. Please stick to obvious facts.
@ednewtonrex The most direct way to tell a cult is that bringing expert objective information to the table is labeled anti-something, ideologized, and so on. Some AI hype and their clueless believers are exactly like that. Ah well.
@samocnj64@jurgen_nauditt Clearly unaware of what psychiatrists use to diagnose histrionic disorder- everyday situations triggering particular textbook behaviours (like in this clip) are key. But he satisfied all clinical criteria already when claimed there was no rain during his 1st inauguration. Bye.
@OunkaOnX This is what typical textboook severe histrionic disorder looks like. Any direct consequence of one's direct actions are dissociated from, actions denied that ever happened, and so on. Any positive actions & consequences by anyone else are appropriated as one's own.
@NoLieWithBTC Some vaguely recally this abomination having been a 200M ballroom. So it already doubled, as expected from this ambulant one-person lunatic asylum.
@MrGrokJr@yazgilbert@Ric_RTP The key is that whatever the nature of its thinking is, it has to show self-consistent reasoning. These LLMs have amply shown absolute inability of doing even the most basic reasoning (see e.g. Stanford study, ARC-AGI-3 benchmark, etc.). So we can get mystical about it but...
@StevenIsserlis The old cliche applies: his music proves there is order in the Universe. Those that stop at "it is mathematical" are like those who look in same way at Gaudi's wire & weights model of La Sagrada Familia cupola and fail to grasp the immense beauty of the building.
@yazgilbert@Ric_RTP Yes... and it is incredibly frustrating to see serious articles & company claims about reducing hallucination. Promises of X% reduction by Y point in time. Etc. They are deliberately (or some, idiotically) walking around the large elephant.
@yazgilbert@Ric_RTP Plus the fun fact that hallucination was directly caused by elementary system design, not AI, problem: they 'forgot' what happens when a system is trained to provide answers, always... instead of signaling that it can't give answer. It is a 1980s lesson... forgotten.
@yazgilbert@Ric_RTP Exactly. It is astounding how they managed to hype around the central aspect that nullifies the big claims. Every truly novel tech had hype around it before organic (!) adoption, but never in the history have so many dangerous idiots push fundamental nonsense via PR & influencers
@yazgilbert@Ric_RTP Plus the fluorescent elephant in the room: when a 4-yr-old has same amount of information assimilated, he/she tasted, touched, seen etc. the physical world. LLMs even if understood anything (they do not) they would have only textual descriptive info as input. Ludicrous.
@Ric_RTP The most worrying thing is that somebody of his caliber has to say this - because, amongst the hype, many have forgotten elementary aspects & lessons from 40+ years old AI research.
@MikeDrucker It is remarkably funny how basic physics (the need for critical mass, and what is needed to achieve that critical mass in specific manner) nullify his insane bullshit. The only funnier thing is how some actually believe & applaud this fundamentally absurd nonsense.